metadata
dataset_info:
features:
- name: audio
struct:
- name: array
sequence: float32
- name: sampling_rate
dtype: int64
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 3243661548
num_examples: 14376
download_size: 3256695722
dataset_size: 3243661548
configs:
- config_name: default
data_files:
- split: train
path: data/train-*